Nowadays internet is an integral part of everyone’s life. However, there are cyber-attackers that interrupt the working of the internet for their gain. Distributed denial of services (DDoS) attacks are one of the popular cyber-attacks that is used by attackers to disrupt the normal working of the internet by exhausting all of the resources of the victim. Due to DDoS, the victim is unable to provide its services to legitimate users, and hence normal work of the internet is interrupted. With the advancement in the field of machine learning and artificial intelligence, researchers are proposing DDoS attack detection techniques based on them. However, due to the diverse nature of DDoS attacks, it is difficult to propose an optimal attack-detection technique. In this context, we proposed a DDoS attack detection technique based on a radial neural network. The proposed approach used the concept of K-means clustering and the radial basis function for the identification of DDoS attack traffic.